← CVS Health Interview Insights

CVS Health·Data Scientist·Technical Phone Screen·Senior

Senior
Jun 2026

Summary

CVS Health data scientist interview that was essentially one long end-to-end experiment design case for a pharmacy flu vaccination campaign. Heavy on stats and causal inference, way more involved than I expected for a single session.

Questions Asked (7)

Q1

You're running a pharmacy flu vaccination campaign over SMS and Email. Would you design this as a simple A/B test or a 2x2 factorial, and why? Walk through your reasoning on interaction effects, interference, and send costs.

A/B Testing & ExperimentationProduct Analytics & MetricsTechnical Trade-offs
Author's notes

I went straight to factorial and felt pretty good about it, but I fumbled the interference piece.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the business goal: to maximize flu vaccination rates while minimizing cost. Then compare a simple A/B test (SMS vs. Email) with a 2x2 factorial (SMS/Email x message variant) by discussing interaction effects, interference, and send costs. Conclude with a recommendation based on whether you expect interactions and whether the incremental cost of the additional cells is justified.

Pro tip: In practice, factorial designs are powerful but can suffer from interference when channels overlap (e.g., patients receiving both SMS and Email). Always consider whether the treatment groups are independent and whether the interaction effect is of primary interest or a nuisance.

1. Clarify the objective and constraints

Define the primary metric (e.g., vaccination rate) and secondary metrics (e.g., cost per vaccination, engagement). Identify constraints such as budget, timeline, and available sample size.

2. Compare design options

Outline a simple A/B test (SMS vs. Email) and a 2x2 factorial (SMS/Email x message variant). Discuss the trade-offs in terms of statistical power, ability to detect interactions, and operational complexity.

3. Analyze interaction effects

Explain that a 2x2 factorial allows testing whether the effect of channel depends on message variant (and vice versa). If no interaction is expected, a simple A/B test may suffice; if interaction is plausible, factorial is preferred.

4. Evaluate interference and send costs

Discuss potential interference: patients may receive both SMS and Email if not properly randomized, leading to contamination. Also, sending multiple messages increases cost; factorial requires more sends, so weigh the cost against the value of interaction insights.

5. Make a recommendation

Based on the above, recommend either a simple A/B test (if resources are limited, no interaction expected, or interference is high) or a 2x2 factorial (if interaction is of interest and costs are manageable). Justify with expected impact on vaccination rates and ROI.

Key Points to Mention

  • Interaction effects: factorial designs can detect synergies or antagonisms between channel and message, which a simple A/B test cannot.
  • Interference: patients might receive multiple messages if not mutually exclusive, biasing results; ensure proper randomization and consider a holdout group.
  • Send costs: factorial requires more messages, increasing cost; calculate incremental cost per additional vaccination to decide if worth it.
  • Statistical power: factorial may have lower power per cell if sample size is fixed; consider overall sample size and expected effect sizes.
  • Business context: CVS Health aims to maximize vaccination rates and minimize cost; the design should align with these goals.
  • Practical execution: consider operational feasibility of implementing multiple message variants and tracking across channels.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Define the eligibility criteria, exclusions, randomization unit, and stratification variables for this experiment.

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

Felt solid here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the business context at CVS Health, then systematically define each component: eligibility criteria, exclusions, randomization unit, and stratification variables. Explain how each choice impacts validity, bias, and statistical power, and tie them to practical constraints like data availability and ethical considerations.

Pro tip: Always align your design with the company's existing data infrastructure and privacy policies; for example, at CVS Health, consider how HIPAA and member consent affect eligibility and data collection. Also, pre-register your analysis plan to avoid p-hacking and demonstrate rigor.

1. Clarify Experiment Objective and Scope

Understand the hypothesis, target population, and success metrics. This ensures all design choices support the experiment's goal and business relevance.

2. Define Eligibility Criteria

Specify who can participate based on demographics, behavior, or tenure. Include criteria that ensure sufficient exposure to the treatment and relevance to the business question.

3. Determine Exclusions

Identify groups to exclude, such as employees, bots, or users with prior exposure, to reduce confounding and interference. Justify each exclusion with data or logic.

4. Choose Randomization Unit

Select the unit of randomization (e.g., user, session, store) based on the experiment's nature and potential spillover. Consider clustering and analysis implications.

5. Select Stratification Variables

Pick variables correlated with the outcome to balance groups and increase power. Use pre-treatment covariates and ensure strata are large enough for stable estimates.

Key Points to Mention

  • Eligibility criteria should be based on the target population and ensure users can be affected by the treatment.
  • Exclusions must be justified to avoid bias and should include employees, bots, and users with prior exposure.
  • Randomization unit choice depends on interference risk and analysis complexity; user-level is common but consider cluster randomization.
  • Stratification variables should be pre-treatment, correlated with the outcome, and have sufficient sample size per stratum.
  • Consider practical constraints like data availability, privacy regulations (e.g., HIPAA), and implementation feasibility.
  • Pre-register the analysis plan to maintain statistical validity and avoid p-hacking.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q3

Compute the per-arm sample size for this experiment given an 8% baseline vaccination rate, 1.5 percentage point MDE, two-sided alpha of 0.05, and 80% power. Then discuss how household clustering with ICC of 0.01 and average household size of 1.3 inflates that number.

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

The formula part was fine, standard two-proportion z-test, came out around 3,500 per arm give or take.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, compute the required sample size per arm for a two-proportion z-test using the given baseline rate, MDE, alpha, and power. Then, adjust for household clustering by calculating the design effect (1 + (average household size - 1) * ICC) and multiply the initial sample size by this factor. Finally, discuss the implications for the experiment, such as increased recruitment needs and potential trade-offs.

Pro tip: Always clarify whether the MDE is absolute or relative; here it's absolute (1.5 percentage points). Also, mention that clustering reduces effective sample size, so you might need to randomize at the household level or use cluster-robust standard errors.

1. Compute base sample size

Use the formula for two-proportion z-test: n = (Z_{α/2} + Z_β)^2 * (p1*(1-p1) + p2*(1-p2)) / (p2-p1)^2, where p1=0.08, p2=0.095 (8% + 1.5pp), Z_{α/2}=1.96, Z_β=0.84. Calculate n per arm.

2. Calculate design effect

Design effect (DE) = 1 + (average cluster size - 1) * ICC. Here, average household size = 1.3, ICC = 0.01, so DE = 1 + (1.3 - 1)*0.01 = 1.003.

3. Adjust sample size for clustering

Multiply the base sample size by the design effect to get the required sample size per arm accounting for clustering. This yields a slight increase (0.3% more).

4. Discuss implications and assumptions

Explain that the inflation is minimal due to small household size and low ICC, but in other scenarios it could be substantial. Mention that if randomization is at household level, the number of households needed is the adjusted sample size divided by average household size.

Key Points to Mention

  • Formula for sample size calculation for two proportions.
  • Definition and calculation of Minimum Detectable Effect (MDE) as absolute difference.
  • Design effect formula for clustered data: 1 + (m - 1) * ICC.
  • Interpretation of ICC and average cluster size in the context of household clustering.
  • Implication of clustering on power and sample size (effective sample size reduction).
  • Potential need for cluster-level randomization or cluster-robust standard errors.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q4

Define the ITT estimator for this campaign. Then use deliverability as an instrument to compute the TOT, show the ITT/compliance relationship, and state the assumptions required.

A/B Testing & ExperimentationProduct Analytics & MetricsTechnical Trade-offs
Author's notes

This is where I slowed down.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the ITT estimator as the difference in means between randomized groups, then introduce deliverability as a binary instrument to compute the TOT via the Wald estimator. Clearly show the ITT/compliance relationship (ITT = TOT × compliance rate) and list the key assumptions (randomization, exclusion restriction, monotonicity, SUTVA) that validate the instrumental variable approach.

Pro tip: Emphasize that ITT is the causal effect of assignment, not treatment, and that deliverability is a valid instrument only if it affects the outcome solely through treatment receipt. Also, mention that in practice, compliance rates are often estimated from the data, and you should check for weak instruments.

1. Define the ITT estimator

State that ITT is the average causal effect of being randomized to the treatment group, regardless of actual treatment received. Formally, ITT = E[Y | Z=1] - E[Y | Z=0], where Z is the randomization indicator.

2. Introduce deliverability as an instrument

Explain that deliverability (e.g., whether the email/message was successfully delivered) is a binary instrument that affects treatment receipt but is independent of potential outcomes given randomization. It must satisfy relevance, exclusion restriction, and monotonicity.

3. Compute the TOT using the Wald estimator

Use the formula TOT = ITT / compliance rate, where compliance rate = E[D | Z=1] - E[D | Z=0] and D is the treatment receipt indicator. This is the local average treatment effect (LATE) for compliers.

4. Show the ITT/compliance relationship

Derive that ITT = TOT × compliance rate, meaning the intention-to-treat effect is diluted by non-compliance. This relationship highlights why ITT underestimates the treatment effect when compliance is imperfect.

5. State the assumptions

List the required assumptions: (1) random assignment (Z independent of potential outcomes), (2) exclusion restriction (Z affects Y only through D), (3) monotonicity (no defiers), and (4) stable unit treatment value assumption (SUTVA). Also note that the instrument must be relevant (compliance rate > 0).

Key Points to Mention

  • ITT estimator definition and formula
  • Deliverability as a binary instrument and its validity conditions
  • Wald estimator for TOT: ITT divided by compliance rate
  • Relationship: ITT = TOT × compliance rate
  • Assumptions: randomization, exclusion restriction, monotonicity, SUTVA
  • Interpretation of TOT as LATE for compliers

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q5

In a factorial design where customers can receive both SMS and Email, how would you instrument and analyze the data to attribute incremental impact to each channel? Why is self-reported attribution biased, and is there any valid use for it?

A/B Testing & ExperimentationProduct Analytics & MetricsRoot Cause Analysis
Author's notes

Proposed unique tracking codes per channel per customer, message timestamps to establish sequence, and then factorial contrasts to estimate main effects and the interaction term.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by describing a 2x2 factorial design with SMS and Email as factors, including a holdout group to measure incremental impact. Explain how to instrument the experiment with randomized assignment and track user-level conversions, then analyze using ANOVA or regression to estimate main effects and interactions. Finally, discuss why self-reported attribution is biased due to recall and social desirability, but note it can be useful for qualitative insights or debugging.

Pro tip: Emphasize the importance of pre-registering the analysis plan and checking for interaction effects, as they can significantly alter channel attribution. Also, mention that self-reported attribution can complement experimental data by providing context on user perception.

1. Design the factorial experiment

Set up a 2x2 factorial design with SMS and Email as factors, including a control group receiving neither. Randomize customers to each cell to ensure unbiased assignment.

2. Instrument data collection

Implement tracking for exposures (SMS/Email sent, delivered, opened) and outcomes (conversions, revenue). Use unique user IDs to link exposures to outcomes and ensure data quality.

3. Analyze incremental impact

Use ANOVA or linear regression to estimate main effects of each channel and their interaction, controlling for covariates. Calculate incremental lift relative to control.

4. Address self-reported attribution bias

Explain biases like recall bias, social desirability, and attribution errors. Discuss how self-reported data can misattribute conversions, especially when multiple channels are present.

5. Identify valid uses of self-reported attribution

Suggest using self-reported data for qualitative insights, hypothesis generation, or debugging tracking issues, but not for causal inference.

Key Points to Mention

  • Randomization and control group are essential for causal inference.
  • Factorial design allows testing of interaction effects between channels.
  • Use of ANOVA or regression to estimate main effects and interactions.
  • Self-reported attribution suffers from recall bias, social desirability, and misattribution.
  • Self-reported data can be useful for qualitative research and hypothesis generation.
  • Ensure proper instrumentation with exposure and outcome tracking at user level.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q6

In one market, 100 customers in the SMS arm got vaccinated but only 50 say they came because of a message. The control vaccination rate is 7.5% and the SMS arm rate is 9.0%. Compute the ITT lift and explain why the self-report number doesn't change your causal estimate. What would you test next if lift falls below 1.5 percentage points?

A/B Testing & ExperimentationProduct Analytics & MetricsRoot Cause Analysis
Author's notes

ITT lift is just 9.0% minus 7.5% = 1.5pp, pretty clean.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by computing the ITT lift as the difference in vaccination rates between the SMS arm and control (9.0% - 7.5% = 1.5 percentage points). Explain that the self-report of 50 customers is subject to attribution bias and does not affect the ITT estimate because ITT measures the causal effect of assignment, not treatment received. Then, outline a plan to test next steps if the lift falls below 1.5 pp, focusing on power, segmentation, and mechanism.

Pro tip: Emphasize that ITT is the gold standard for causal inference in randomized experiments because it preserves randomization and avoids selection bias. Mention that self-reports are often unreliable due to social desirability and recall bias, so they should be used for exploratory insights, not causal claims.

1. Compute ITT Lift

Calculate the difference in vaccination rates between the SMS arm and control: 9.0% - 7.5% = 1.5 percentage points. This is the intent-to-treat (ITT) effect.

2. Explain Why Self-Report Doesn't Change Causal Estimate

The self-report (50 customers) is a subset of the 100 vaccinated in the SMS arm and is subject to attribution bias. ITT measures the effect of being assigned to the SMS arm, regardless of whether customers saw or acted on the message, so it remains valid.

3. Address If Lift Falls Below 1.5 pp

If the lift is below 1.5 pp, first check statistical significance and power. Then, explore heterogeneity by customer segments, message timing, and frequency. Consider whether the control group had contamination or if the SMS delivery failed.

4. Propose Next Tests

Suggest A/B tests with different message content, send times, or incentives. Also, test alternative channels (e.g., email, app notifications) and measure incremental lift via holdout groups.

Key Points to Mention

  • ITT vs. treatment-on-the-treated (TOT) and why ITT is preferred in randomized experiments.
  • Self-report bias: social desirability, recall bias, and attribution error.
  • Statistical significance: compute confidence intervals and p-value for the 1.5 pp lift.
  • Power analysis: ensure the experiment is adequately powered to detect a 1.5 pp difference.
  • Segmentation: analyze subgroups (e.g., age, prior vaccination status) to find heterogeneous effects.
  • Mechanism: consider whether the SMS was delivered, opened, and acted upon; use engagement metrics as mediators.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q7

How would you monitor the experiment during rollout, finalize results, and recommend a scaled channel policy under a fixed budget?

A/B Testing & ExperimentationProduct Analytics & MetricsProduct Strategy
Author's notes

Talked through sequential testing with alpha spending to allow early stopping without inflating type I error.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around a clear experimentation lifecycle: monitoring, analysis, and decision-making. Emphasize statistical rigor, business impact, and budget constraints. Use a concrete example to illustrate how you would handle each phase.

Pro tip: Show that you balance statistical significance with practical significance—especially in healthcare where small lifts can have large financial or clinical impact. Mention how you would handle multiple comparisons and guardrail metrics to avoid false positives.

1. Define success metrics and guardrails

Clearly state primary, secondary, and guardrail metrics (e.g., conversion, revenue, customer satisfaction, safety). Establish minimum detectable effect and required sample size to ensure adequate power.

2. Monitor experiment health and early signals

Set up real-time dashboards to track sample ratio mismatch, metric trends, and guardrail violations. Use sequential testing or alpha spending to allow early stopping for harm or futility without inflating Type I error.

3. Analyze results with statistical rigor

After the experiment concludes, perform hypothesis tests (e.g., t-test, bootstrap) and calculate confidence intervals. Adjust for multiple comparisons if needed, and check for heterogeneous treatment effects across key segments.

4. Translate results into business impact and budget trade-offs

Estimate the incremental lift and its monetary value. Compare against implementation costs and budget constraints. Consider scalability, operational feasibility, and potential risks.

5. Recommend a scaled channel policy

Propose a rollout plan (e.g., full scale, phased, or targeted) based on ROI and budget. Define monitoring and re-evaluation cadence, and suggest a decision framework for future experiments under fixed budget.

Key Points to Mention

  • Statistical power, sample size calculation, and minimum detectable effect
  • Sequential testing or alpha spending for safe interim monitoring
  • Guardrail metrics and safety checks, especially in healthcare
  • Heterogeneous treatment effects and segment analysis
  • ROI calculation and budget allocation optimization
  • Scalability, operational constraints, and phased rollout strategies

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.